Econometric methods for the analysis of banking data

This course provides tools to analyse the properties of disaggregated banking data, their cyclical characteristics and to test hypotheses regarding the behaviour of banks in the funding and lending markets. It studies situations where bank heterogeneities are constant and when heterogeneities evolve over time and when they affect both the level and the dynamics of their lending activities. It also provides some explanations for some empirical puzzles found in the literature.

Available methods have been drawn from the microeconomic literature where a large cross section and a small time series are typically available. Banking data typically have a limited number of cross-sectional units of medium term length. Thus, standard asymptotic results do not apply. In addition, while standard methods can deal with level or variance heterogeneities, they have difficulties to deal with slope heterogeneities which are very relevant to study pass-through and risk features. To properly analyse panels of banking data, alternative techniques and ideas are needed. This course provides both.